1.0 Nonprecedential Opinions Impede the Law's Development

A. Reality

Since at least the 1980s but before Federal Rules of Appellate Procedure rule 32.1 became effective January 2007, federal circuits enforced a bizarre rule subjecting attorneys to monetary sanctions for citing unpublished local-circuit opinions. What was strangest about this common rule was its coexistence with rules allowing citation to unpublished opinions authored by other circuits. Why are the circuits so modest in estimating the relative worth of their own unpublished opinions, conferring less authority on them than opinions from other circuits, less authority than even a student’s law-review comment? Why would the courts go further than banishing these opinions from consideration, adding punishment for their mere mention, when striking a circuit's disowned progeny suffices? These unpublished opinions were disparaged more than were irrelevant facts or foundationless claims.

Judges, occasionally, have been candid enough to reveal why they are averse to citing their unpublished opinions: the judges are uncertain the unpublished opinions were correctly decided and embarrassed by possible contradictions with precedential cases. Judge Kozinski of the Ninth Circuit has become the main representative for the judiciary's view — that the nonprecedential opinions are merely less well articulated, resembling more a letter of explanation to the parties concerned than a rigorous and complete justification. (See Hart v. Massanari (9th Cir. 2001) 266 F.3d 1155.) This “official” account doesn't help explain the judiciary's phobia of having these opinions cited in their home courts. The judiciary's official view ignores that the practice of treating opinions as precedential only when the court wants them so treated eliminates half the impetus to correct judgment. Lawyers understand stare decisis protects litigants by adding an additional constraint on the judge's discretion. But stare decisis provides another form of protection. Courts are constrained not only by having to conform their conclusions to what happened previously but by the knowledge that their conclusions control future cases.

B. Legality

Since the late Judge Richard Arnold wrote Anastasoff v. U.S. (8th Cir., 2000) 235 F.3d 154 [vacated for other reasons], declaring the practice of nonprecedential opinions violates Article III, section 2, of the U.S. Constitution, academics argue whether nonprecedential opinions are constitutional. Counterposed to Anastasoff is the 9th Circuit's rebuttal, Hart v. Massanari, supra, 266 F.3d 1155. The 8th Circuit panel concluded that the Founders had conceived of the judicial power as creating law with each appellate case. Judge Kozinski countered with scholarship showing that the practice of citing specific cases as precedent developed only after casebooks became available in the eighteenth and early nineteenth centuries, earlier courts having embraced a less articulated doctrine of decision according to existing law. Judge Kozinski's argument confuses a procedural history where stare decisis is invoked without citing any specific case with one where the court bans a case’s precedential use. Judge Kozinski also argues that if precedent were inherent in the Framers' concept of the judicial power, it would also apply to the trial courts, but an argument based on the denial of precedential significance to trial court judgments ignores juries’ centrality for the Framers. Arguments such as Judge Kozinski's lead even those sympathetic to Judge Arnold to criticize his scholarship for insufficiently specific evidence on the Founders' intent. (See, for example, Eighth Circuit Holds That Unpublished Opinions Must Be Accorded Precedential Effect (2001) 114 Harv. L. Rev. 940.) The criticism misconceives Arnold's argument, which is based on the original constitutional concept of judicial power, not the details of its execution. The basic question is what differentiates the judicial power from the legislative power under the U.S. Constitution, to which Judge Arnold replied the Framers conceived the distinction turning on stare decisis.

A similar argument, with the advantage of its not being subject to historical deflection, can be based on the Article III, section 2, cases and controversies clause. In making a case precedential, the court does not decide a case or controversy. Instead, it chooses among the resolutions which case will serve as precedent, its act of choice outside any case resolution. Probably the reason Judge Arnold didn't make this more constitutionally precise argument is that the courts do other things besides resolving cases and controversies, specifically, they establish rules for resolving cases and controversies. The 8^th Circuit’s analysis in terms of traditional judicial power incorporates the court's rulemaking power, whereas using the cases and controversies clause requires distinguishing among rules so that the nonpublication rule is a different kind than a page-limit rule.

The distinction between actions that do or do not abridge substantive rights is apropos but perhaps ad hoc in justifying an interpretation of a limitation to cases and controversies. (See Shannon, May Stare Decisis Be Abrogated by Rule? (2006) 67 Ohio State L.J. 645, 666 ["the primary purpose of such a rule is to limit the effect of a court's decisions (and therefore the extent and nature of the substantive law) in future cases"].) The cases and controversies clause, after all, has generally been used as a limitation on standing, not as a limit on the kinds of activities the court can engage. But if the practice of making some opinions nonprecedential conflicts with Article III, it conflicts because the court does something besides deciding a case. This argument from the cases and controversies clause requires that its proponents distinguish the rulemaking exceptions as constitutionally inert. The reader can compare the cogency of my argument that pure procedural matters fall outside the U.S. Constitution compared with Judge Arnold's that historical deviations from modern concepts of precedent are surplusage to basic stare-decisis doctrine. (But see Pather, Outing the Judicial Epistemology of Hart V. Massanari (2005) 62 Wash. & Lee L. Rev. 1553 ["The conclusion is that in Hart, Judge Kozinski is asking us to take a lot on trust, and that the record invites a decision against him"].)

C. Policy

In light of the judiciary’s opinions on a question where professional identity usually determines attitude, the constitutionality of nonprecedential opinions is not a live question. If almost all judges favor nonprecedential opinions, academics are even closer to unanimity in opposing them, and practicing attorneys, though speaking less with one voice, are strongly against them. Judges favor nonprecedential opinions because the courts invented this practice to handle the press of growing caseloads. But if the number of judges is insufficient, the most principled solution, as Judge Arnold pointed out in Anastasoff v. U.S., supra, is to create more judgeships. Barring that solution, the proper alternative is to give each case as much attention as it requires, and tolerate the consequent backlog. At the time, some analysts estimated that the number of federal circuits would have to be tripled for publishable opinion quality in each case. The number of cases had expanded because of an expansion of rights during the 1960s, legislatively as well as judicially created, and the number of judgeships didn't keep pace during the decades of economic retrenchment starting in the 1970s. The expansion occurred in two main areas: integration cases deriving from Brown v. Bd. of Educ. and prisoner rights cases under the Civil Rights Act. Evidence suggests that as early as the 1950s, the conservative Southern Fourth Circuit began creating a separate category of unpublished cases to resist Supreme Court initiatives. (See Pether, Inequitable Injunction: The Scandal of Private Judging in the U.S. Courts (2004) 56 Stan. L. Rev. 1435.) Later on, cases favoring socially privileged interests showed a more favorable outcome for those interests among the unpublished cases, suggesting that judges decided the legally ineffectual unpublished cases according to politics more than law. This bias, however, does not mean, as Penelope Pether implies (ibid.), that treating all cases as precedential would produce better results for the powerless, as other subterfuges might have the same discriminatory consequences.

How much accuracy judges lose by freeing themselves of accountability to future cases — not only where negligent or deliberate discrimination is involved —depends on circumstances. Where the potential penalty for failing to follow precedent is reversal, the main potential penalty for failing to create a workable rule is having the case overruled, a much rarer event than reversal, even before the advent of nonprecedential opinions. The lost dimension when courts abjure precedent concerns not quality of decision but quality of law, which results from considering a pool of cases eight or nine times larger, were all judgments treated as precedents. Commentators have characterized a system of nonprecedential opinions as amounting to moving from a common law to a civil law system, and the common law versus civil law framework can help establish a balance sheet for rules permitting nonprecedential judgments. In the civil law, codes rather than precedents determine outcomes, and the nonprecedent system resembles civil law — to be exact, the courts resemble the legislature running a civil-law system — in that the courts sacrifice their judicial law-creation function to perform a legislative function, creating law apart from any case or controversy.

There may be no general answer to whether the common law or civil law system is superior in its decisional output, but each system carries specific advantages. With competent, diligent, and honest judges, the civil law system is capable of delivering better decisions on issues of law. As Frederick Schauer shows in Precedent, a precedent-based system's decisions necessarily trade optimality for consistency to the extent precedent determines case outcome. (See Schauer, Precedent (1987) 39 Stan. L. Rev. 571.) Schauer doesn't recognize, however, that precedent also fosters accuracy when judging quality is further from ideal. Consistency and accuracy are sometimes cooperative outcomes, as a precedential standard leads to greater consistency because of its greater ease of application, achieved by reducing the factors the decider of fact considers. On their terms, precedents will be more correctly applied than law unsupplemented by precedent, and when the gap between optimum and achievement is great enough, the precedent system will prove superior. If the right way to think about the quality problem were conceiving the idealized judge — competent and conscientious —then a nonprecedent system would produce juster outcomes than a system with precedents appended to guide judgment. The appended precedents make the standard easier to apply by giving it an operational definition but harmfully limit a judge who doesn't need the operationalization.

An issue separate from the quality of decision given the same basic corpus of law is the quality of the standards themselves. Under a common law system, the law develops in response to each appellate case and benefits from resolving conflicts between courts through further appeal, whereas under a civil law system, the individual cases don't determine the law, which remains the same after a case is decided. Case by case, the civil law only assimilates the facts, never accommodating to them. Yet, the civil law does not stand apart from any dialectic of legal theory. While civilian judges don’t contribute to the civil law's evolution, "jurist" academics debate and revise the legal codes. The law must develop not only to keep up with the times but to evolve higher levels of fairness and effectiveness. The civil law's development isn't neglected but entrusted to esteemed academics instead of judges.

A regime of nonprecedential decisions jettisons the common-law mechanism of legal evolution. Judges decide among potential precedents which ones should be the real precedents not by deciding cases — that is, not by acting as judges —but by making generalized judgments that try to second-guess the law's optimal development—acting as bureaucrats or legislators. But they do this lacking the broad base of knowledge characteristic of a modern legislature and, most important in comparing to the civil law, lacking the advice of jurists, as common-law commentators play a minor role in drafting statutes and deciding precedential cases. The common law and civil law use different methods of evolving law; in civil law, conflict of ideas and legal artisanship are factored into the system at a different point, in a more philosophically rationalist fashion. Decisions without precedential effect in a common-law system cripple its main vehicle for the law's development without replacement by a civilian-style juristic alternative. The main toll of nonprecedential decisions, ignored by commentators, lies not as much in applying existing law poorly as retarding law's development.